FUELS AND CHEMICALS FROM THE PYROLYSIS OF SCRAP TYRE: OPTIMISATION USING RESPONSE SURFACE METHODOLOGY AND ARTIFICIAL NEURAL NETWORK

AZETA, OSARHIEMHEN and Covenant University, Theses (2021) FUELS AND CHEMICALS FROM THE PYROLYSIS OF SCRAP TYRE: OPTIMISATION USING RESPONSE SURFACE METHODOLOGY AND ARTIFICIAL NEURAL NETWORK. Masters thesis, Covenant University.

PDF Download (392kB)

Abstract

Scrap tyres generated are indiscriminately disposed without consideration for their impacts on human health and the environment. Their non-biodegradable materials constitute a major challenge in the environment. High volatile matter and fixed carbon content make their disposal a cumbersome task. Pyrolysis of automobile scrap tyres was investigated in this study while paying attention to variation and optimisation of the process parameters for the best product yields. Response surface methodology (RSM) was adopted for the optimisation of process variables and development of a statistical model after initial determination of the experimental design runs based on the Box-Behnken design (BBD) approach. Artificial neural network (ANN) modelling was used to predict the accuracy of models obtained from the RSM. Process parameters; residence time (40, 50 and 60 min), temperature (450, 500, 550 °C) and particle size (6.3 mm, 9.4 mm, 12.5 mm) were used for the experimental design. The optimised conditions were validated via thermal pyrolysis and a variation using catalytic pyrolysis with zinc chloride as the catalyst. A fixed bed reactor was utilised for this purpose with a water reservoir connected to the condenser for efficient cooling. The impact of emitted gases on the operator and the surrounding, effect of pyrolysis time, temperature and feed particle size on the pyro-oil produced were assessed as well as the characterisation of the pyro-oil and char. An ANN model based on feed-forward learning algorithm was trained, validated and tested using experimental data points obtained from the RSM in the ratio 70:15:15 respectively to give regression coefficients (R) values for the product yields. The resultant yields of 31.89 wt.% and 37.10 wt.% were obtained for the thermal and catalytic pyrolysis at optimised conditions of pyro-oil respectively at operating time, 60 min, temperature of 503 °C and feed particle size of 6.3 mm at a heating rate of 7 °C/min. The RSM and ANN techniques were proven to be effective tools in the generation of models for the optimisation of pyro-oil yield and can serve as an alternative for laboratory study having R2 values with high degrees of accuracy of RSM (0.9985) and ANN (1.000). Also, the model equations derived from RSM were statistically significant having P-value < 0.05, large F-value (735.76) and optimal composite desirability factor of 0.9793. Fuel properties of the derived pyro-oil were analysed and found to be suitable for use as liquid fuel based on minimal sulphur content of 0.07 – 0.22 %, excellent viscosity (2.93 – 3.36 cSt), density (0.889 – 0.918 g/cm3) and higher heating values of 35.40 – 44.24 MJ/kg. A detailed characterisation of the pyro-oil was performed using FTIR, and GC-MS while BET, SEMEDX, XRF and XRD were performed on the char. Based on the analyses carried out, it can be said that the pyro-oil being a complex mixture of organic compounds can serve as feedstock for industrial processes. Also, the properties based on the physiochemical properties encourage the use of this oil and char as conventional fuels, fillers and pigments.

Item Type:	Thesis (Masters)
Uncontrolled Keywords:	Pyrolysis, Automobile Scrap tyre, Pyrolytic oil, Char, Response surface methodology, Artificial neural network
Subjects:	T Technology > T Technology (General) T Technology > TP Chemical technology
Divisions:	Faculty of Engineering, Science and Mathematics > School of Engineering Sciences
Depositing User:	Mrs Hannah Akinwumi
Date Deposited:	01 Jul 2022 11:03
Last Modified:	01 Jul 2022 11:03
URI:	http://eprints.covenantuniversity.edu.ng/id/eprint/16047

Actions (login required)

View Item